Altered cortisol metabolism in polycystic ovary syndrome: insulin enhances 5alpha-reduction but not the elevated adrenal steroid production rates.

Androgen excess in women with polycystic ovary syndrome (PCOS) may be ovarian and/or adrenal in origin, and one proposed contributing mechanism is altered cortisol metabolism. Increased peripheral metabolism of cortisol may occur by enhanced inactivation of cortisol by 5alpha-reductase (5alpha-R) or impaired reactivation of cortisol from cortisone by 11beta-hydroxysteroid dehydrogenase type 1 (11beta-HSD1) resulting in decreased negative feedback suppression of ACTH secretion maintaining normal plasma cortisol concentrations at the expense of androgen excess. We have tested whether any enzyme dysregulation was related to circulating insulin or androgen concentrations in women with PCOS and have sought to clarify their relationship with obesity. First, to avoid obesity-related effects on cortisol metabolism, 18 lean women with PCOS were compared with 19 lean controls who were closely matched for body mass index (BMI). Second, the impact of obesity was studied in a cross-section of 42 PCOS women of a broad range of BMI. We measured 24-h urinary excretion of steroid metabolites by gas chromatography/mass spectrometry and fasting metabolic and hormone profiles. Urinary excretion of androgens [androsterone (P = 0.003), etiocholanolone (P = 0.02), and C19 steroid sulfates (P = 0.009)], cortisone metabolites [tetrahydrocortisone (THE) (P = 0.02), alpha-cortolone (P < 0.001), beta-cortol + beta-cortolone (P < 0.001), cortolones (P < 0.001), and E metabolites (P < 0.001)], and TCM (P = 0.002) were raised in lean PCOS subjects when compared with controls. A significantly higher 5alpha-tetrahydrocortisol (5alpha-THF)/5beta-THF ratio (P = 0.04) and a significantly lower alpha-THF + THF + alpha-cortol/THE + cortolones ratio (P = 0.01) were found in lean PCOS women compared with lean controls, indicating both enhanced 5alpha-R and reduced 11beta-HSD1 activities. A decreased THE/cortolones ratio (P = 0.03) was also found in lean PCOS women compared with lean controls, indicating increased 20 alpha/beta-HSD activity. In the group of 42 PCOS subjects, measures of 5alpha/5beta reduction were positively correlated with the homeostasis model insulin resistance index (HOMA-R): alpha-THF/THF and HOMA-R (r = 0.34; P = 0.03), androsterone/etiocholanolone and HOMA-R (r = 0.32; P = 0.04), and total 5alpha /total 5beta and HOMA-R (r = 0.37; P = 0.02). A positive correlation was also found between measures of 5alpha-R and BMI (r = 0.37; P = 0.02). No correlation was found between measures of 11beta-HSD1 activity and indices of insulin sensitivity or BMI. We have demonstrated that there is an increased production rate of cortisol and androgens as measured in vivo in lean PCOS women. Insulin seems to enhance 5alpha reduction of steroids in PCOS but was not associated with the elevated cortisol production rate. The changes in 5alpha-R, 11beta-HSD1, and 20alpha/beta-HSD enzyme activities observed in PCOS may contribute to the increased production rates of cortisol and androgens, supporting the concept of a widespread dysregulation of steroid metabolism. This dysregulation does not seem to be the primary cause of PCOS because no correlation was found between serum androgen levels or urinary excretion of androgens with measurements of either 5alpha-R or 11beta-HSD1 activities.